Features
Real-time video generation at 25 fps
Keyboard and mouse input control
Auto-regressive diffusion-based image-to-world framework
Long video generation with temporal coherence
Open-source code and models (MIT license)
Three pretrained scenes: universal, GTA driving, TempleRun
Demo utilities and configuration files
Hugging Face model integration
Local inference on Linux with NVIDIA GPU
Generate video with Hedra Character-3 model
Generate talking-head avatar videos from photo and audio
Generate cinematic clips using Veo 3.1, Kling 3, Seedance, MiniMax, Sora 2 Pro
Generate images with Flux, Imagen4, Ideogram, Seedream 4.5, Nano Banana
Generate audio with ElevenLabs, Cartesia, MiniMax
Hedra Agent for conversational content creation (intent to finished content)
Reusable Skills for workflow automation (save and reuse custom workflows)
Agent learns from user's data and reuses brand elements
Search the web during creative process
Infinite canvas for multi-modal generation
Connect your own data sources
Lip-sync avatar videos with voice cloning (4000+ voices)
Camera angle control via prompts
Up to 10-minute avatar video generation
Teams collaboration and SSO (Enterprise)
